Omnidirection image restoration using a Support Vector Machine

Omni-directional vision produces a spherical field of view of an environment, which appears definite significant since its advantage of panoramic sight with a single compact visual scene. The Omni-directional vision is provided by various image systems. Fisheye lens is one of the most efficient ways to establish omni-directional image systems. However fisheye lens images, which can include a 2π steradian field of view, appear with an unavoidable inherent distortion. But it can be corrected with image processing techniques. A method for geometric restoration of such distorted images is derived in this paper. A simple method of correcting the distortion of fisheye image by means of a Support Vector Machine (SVM) to replace ordinary correction model is proposed., SVM is a machine learning method based on the theory of statistics, which have good capabilities of imitating, regression and classification. The approach using SVM provides a mapping between the fisheye image and the standard image for human eyes, which involves a coordinate conversion between fisheye image and real world view. According to this method we don’t need to evaluate the various parameters of the distortion and to concern the projection model of fisheye lens which usually needs to be acquired from the manufacturer. Finally, the experiments demonstrate the feasibility of the method for real time image process and the results are satisfactory for image restoration.